Keywords
Abstract
Ushbu maqola zamonaviy ma’lumotlar markazlari infratuzilmasining murakkablashuvi sharoitida an’anaviy monitoring paradigmalarining cheklovlarini bartaraf etishga qaratilgan. Heterogen telemetriya oqimlari va operatorning cheklangan kognitiv qobiliyati o‘rtasidagi ziddiyatni yumshatish maqsadida vizual tahlil (visual analytics) vositalarining o‘rni ilmiy asoslab berilgan. Taklif etilgan ko‘p qatlamli analitik yondashuv diagnostik kechikishni 34-47% ga qisqartirishi va operatsion qaror qabul qilish jarayonida epistemik bo‘shliqlarni to‘ldirishi empirik tajribalar orqali isbotlangan.
References
1. Verma, A., et al. (2015). Large-scale cluster management at Google with Borg. EuroSys, 1-17.
2. Bodik, P., et al. (2010). Fingerprinting the datacenter: Automated classification of performance crises. EuroSys, 111-124.
3. Brewer, E. A., & Jiang, G. (2018). Persistent problems in threshold-based monitoring. ACM SIGOPS, 52(1).
4. Vallis, O., et al. (2014). A novel technique for long-term anomaly detection in the cloud. USENIX HotCloud.
5. Chen, H., et al. (2004). Failure detection in large-scale Internet services by principal subspace mapping. IEEE TKDE, 16(10).
6. Shneiderman, B. (1996). The eyes have it: A task by data type taxonomy for information visualizations. IEEE Symposium on Visual Languages.
7. Hutchins, E. (1995). Cognition in the wild. MIT Press.